gbrown@tybalt.caltech.edu (Glenn C. Brown) (03/15/90)
It must be really frustrating trying to come up w/ a HDTV standard: I mean these guys (the ones making the standard) have to come up with the LAST WORD in TV standards. The standard has to be something we're willing to live with for AT LEAST the next 50 years! Sure, 50 years from now, the things will be cheap to make, but now they're going to be VERY expensive. Especially if they make no compromises so we'll still be happy with the standard in 50 years... But you know all that. I'd be perfectly happy to settle for the telivision picture tubes of today! It's the signal that's so horrendous! I mean: If you've ever seen the output of a laser disc player, it's awesome! It shows what your picture tube can do. It's the low signal to noise ratio of the broadcast signals that we receive that's horrendous! I commend the FCC for requiring that the new format signals be backwards compatible, but I think that the standard could easily offer more than just backwards compatibility. Here's what I mean: +----+--------------+----+ I've heard of one standard which would break | | | | your picture into at least four signals: First | A | B/C | D | a standard TV picture is sent (B). This provides | | | | the requisite backwards compatibility. Then | | | | image C is sent. This image is interlaced between | | | | the lines from picture B. Third, zones A&D are +----+--------------+----+ sent to provide HDTC users w/ a movie-box picture. If a standard such as this were adopted (which I highly doubt it would), there would be an EASILY EXTRACTED HD IMAGE OF THE NORMAL PICTURE. Ok, so it's not exactly the same image (It's the line in-between the normal NTSC lines), but it could easily be extracted. And if images A,C,&D are sent digitally w/ error correction, a (relatively) cheap tuner could be built to extract the high-quality image C, convert it to analog, and display this VIRTUALLY NOISE-FREE picture on your old TV! This way, consumers would have a choice of three levels of TV quality, depending upon what they could afford: They could use the cheap old NTSC, or the crisp new digital NTSC, or the dramatic full-performance HDTV! I personally don't think the above standard will (or even should) be implemented as I have described it, But I highly encourage the HDTV Standards Committee to use a format which includes an easily tuned digital encoding of the current NTSC signal. --Glenn
billd@fps.com (Bill Davidson) (03/16/90)
In article <1990Mar15.090214.9871@spectre.ccsf.caltech.edu> gbrown@tybalt.caltech.edu (Glenn C. Brown) writes: > I'd be perfectly happy to settle for the telivision picture tubes of today! >It's the signal that's so horrendous! I mean: If you've ever seen the >output of a laser disc player, it's awesome! It shows what your picture tube >can do. It's the low signal to noise ratio of the broadcast signals that we >receive that's horrendous! I disagree completely. Resolution is sickly compared to what it could be. Have you ever seen a high resolution screen? Laser disks are as good as it gets on a regular TV but the resoulution is still pitiful. The pixels are huge and there's no signal that can possibly fix that. Maybe I'm spoiled by looking at 1000+ line computer screens so much (I'd like HDTV to be even higher resolution but that would *really* be expensive). > I commend the FCC for requiring that the new format signals be backwards >compatible, but I think that the standard could easily offer more than just >backwards compatibility. [idea deleted --billd] How long do we have to carry around the baggage of a standard that was designed so long ago that it can't even get the colors right most of the time? Color was an add-on and the implementation suffered in order to maintain compatibility with old black and white sets. At some point you have to say "enough is enough". We can do so much better now. We know a lot more about video signals than we did when NTSC was designed. Also, the frame rate is annoying. It destroys resolution when converting 24 frame/sec film to video due to frame mixing. I want a standard with at least 1000 lines and a 72Hz frame rate. Wide screen would be nice for films and square pixels would be nice for computer graphics. Why suffer with the old forever? Just because most people won't be able to afford it is rediculous. Most people couldn't afford pocket calculators when they first came out (or TV's, or cars or most other major new technologies). We need to define a standard that is good and is doable and which can be foreseen to become cheap with time. It doesn't have to be cheap now. It would be nice if it was an international standard as well so that video tapes and laserdisks will work anywhere. It took CD's 5-6 years to really break into the US market. A lot of people thought they were rediculous when they first became available. They were very expensive (both the players and the disks). Now it's getting hard to find records stores that have more vinyl than aluminum coated plastic. It took laser video even longer (it's back and gaining a lot of momentum right now). HDTV will be the same story. We have generations of people now who grew up watching TV and they are getting more and more demanding of quality video. --Bill
bill@bilver.UUCP (Bill Vermillion) (03/16/90)
In article <1990Mar15.090214.9871@spectre.ccsf.caltech.edu> gbrown@tybalt.caltech.edu (Glenn C. Brown) writes: > It must be really frustrating trying to come up w/ a HDTV standard: I mean >these guys (the ones making the standard) have to come up with the LAST WORD >in TV standards. The standard has to be something we're willing to live with >for AT LEAST the next 50 years! Sure, 50 years from now, the things will be >cheap to make, but now they're going to be VERY expensive. Especially if they >make no compromises so we'll still be happy with the standard in 50 years... We've already had current video standards for 50 years (though the color portions a bit less than that) and it's time for a change. If you consider that the LP is virtually dead now, it only lasted 41. > I'd be perfectly happy to settle for the telivision picture tubes of today! >It's the signal that's so horrendous! I mean: If you've ever seen the >output of a laser disc player, it's awesome! It shows what your picture tube >can do. It's the low signal to noise ratio of the broadcast signals that we >receive that's horrendous! The output of a laser player isn't awesome. It's good, but you can surely see the limits of NTCS video if you watch it on anything bigger than a 21" set. I have had a laser player for several years now. The early discs are poor compared to today's technology, but we are pretty much at the limits. I assume you are a recent disc convert. S/N is only one problem. Have you ever noticed how the color is in little dots at the transition points, how the resolution isn't as good as a clean 16mm print, let alone 35? > I commend the FCC for requiring that the new format signals be backwards >compatible, but I think that the standard could easily offer more than just >backwards compatibility. ..... I spent years in broadcast, and have seen many changes in the FCC, and I am not too particularly impressed with their performance in the past few years. They seem to be a non-regulatory regulatory agency. Part of the time it's hands off, other times its hands on for the wrong reasons. They totally blew the AM stereo standards by refusing to take a stand. Years ago, here in Orlando, a disk jockey "locked" himself in a radio station control room and played Sheb Wooley's "Monkey Fever" for 24 hours straight. His name was Mark Fowler. He didn't do much better when he was chairman of the FCC. -- Bill Vermillion - UUCP: uunet!tarpit!bilver!bill : bill@bilver.UUCP
news@haddock.ima.isc.com (overhead) (03/16/90)
In article <530@bilver.UUCP> bill@bilver.UUCP (Bill Vermillion) writes: >If you consider that the LP is virtually dead now, it only lasted 41. I never liked LPs. Even an audiophile LP has pops & klicks, even the first play. It gets worse each play. A low grade audio cassette doesn't have terrible distractions or sound degradation. We could have dumped LPs long ago. The industry kept pushing vinal. They kept saying that it was "better". A friend has a nice 25 inch monitor TV. It was real expensive. I saw some stuff on laser disk. It was very impressive. I saw some stuff on video tape - super beta, VHS. Less impressive, but not generally distracting. Cable TV had almost OK stations and pretty bad ones. Without cable, the distractions while watching the show are almost as bad as commercials. If over-the-air quality could be brought to laser disk standards it would be an overnight success. If it was cable-only, I'd probably buy a better TV & spend the $30-$40 a month. Heck, I might even watch it. If HDTV doesn't have some sort of correction system built in, then it will be no better than cable as it is now. Simply increasing bandwidth on a noisy medium does not remove noise. If you create un-expandable standards, they will be bad. The rope cut to length is too short. Technology will outstrip our current ideas of what is easily "as much as anyone can afford". If you want to build a cheap TV with the new standard, you should be able to ignore things - like error correction, or some of the resolution, or one of the sound channels, or the closed captioned channel... On transmission, you have to be able to omit stuff, like error correction, resolution, sound channels... Stephen. suitti@haddock.ima.isc.com
bowers@elxsi.dfrf.nasa.gov (Al Bowers) (03/17/90)
In article <530@bilver.UUCP> bill@bilver.UUCP (Bill Vermillion) writes: >In article <1990Mar15.090214.9871@spectre.ccsf.caltech.edu> gbrown@tybalt.caltech.edu (Glenn C. Brown) writes: ...stuff deleted... >> I'd be perfectly happy to settle for the telivision picture tubes of today! >>It's the signal that's so horrendous! I mean: If you've ever seen the >>output of a laser disc player, it's awesome! It shows what your picture tube >>can do. It's the low signal to noise ratio of the broadcast signals that we >>receive that's horrendous! >The output of a laser player isn't awesome. It's good, but you can surely see >the limits of NTCS video if you watch it on anything bigger than a 21" set. I >have had a laser player for several years now. The early discs are poor >compared to today's technology, but we are pretty much at the limits. I >assume you are a recent disc convert. S/N is only one problem. Have you >ever noticed how the color is in little dots at the transition points, how the >resolution isn't as good as a clean 16mm print, let alone 35? Just as a point of reference laser disc has always been a superior to tape. The fundamental reason for this is the digital nature of LDs. Keep in mind that a good VHS resolution from a decade ago was 190 lines of horizontal resolution, improvments in signal conditioning and processing that apply to the analog tape formats will apply just as well to the analog portion of a LD signal once it has run through a DAC (digital to analog converter). Photo rag tests have shown that lenses for film are capable of up to 100 lines per millimeter resolution (the test I recall said a Lietz 50mm f/2.0 could get 102 l/mm in the center of the frame). On the basis of that I'd say 35mm is capable of in the range of 20000 lines resolution across a frame from edge to edge. 16mm is probably half of that. The real limit in film is the films ability to resolve, probably in reality half of the above numbers, 10000 lines fo 35mm and 5000 lines for 16mm. So I'd say we still have an order of magnitude to go yet. But give me a clean signal and I can live with a lot less resolution. S/N is where video can really beat out film, especially if its digital. I'd vote for a clean break from the current NTSC standard in the new HDTV standard. Give me anywhere from 500 to 1000 lines horizontal resolution and I could be happy with that until I'm old and gray... Just another opinion... -- Albion H. Bowers bowers@elxsi.dfrf.nasa.gov ames!elxsi.dfrf.nasa.gov!bowers NASA Ames-Dryden Flight Research Facility, Edwards, CA Aerodynamics: The ONLY way to fly! Live to ski, ski to live...
bas+@andrew.cmu.edu (Bruce Sherwood) (03/17/90)
At the risk of stating the obvious: Some of this discussion of HDTV implies that there will always be a need for another generation of standards with even higher resolution. That isn't necessarily the case. The human eye has limited resolution, and higher resolution than that in the picture is literally useless, if you are talking in terms of a "typical" screen size viewed from a "typical" viewing distance. Similarly, there must be an upper limit on useful fidelity in color discrimination, beyond which the human eye just can't see any improvement. The analogy with audio is that a CD with frequency response out to 10 MHz would not sound better than one with frequency response out to 20 KHz, because the human ear can't hear the higher frequencies. What we want in electronic products is high fidelity for both eye and ear, but no more than that. Unfortunately today's television and computers are typically well below this threshold. Bruce Sherwood
kucharsk@number6.Solbourne.COM (William Kucharski) (03/17/90)
In article <BOWERS.90Mar16105816@drynix.dfrf.nasa.gov> bowers@elxsi.dfrf.nasa.gov (Al Bowers) writes: >Just as a point of reference laser disc has always been a superior to >tape. The fundamental reason for this is the digital nature of LDs... Sorry, but the video portion of laser videodiscs is analog; the only portion which may be digital on any given LD is the audio... -- =============================================================================== | ARPA: kucharsk@Solbourne.COM | William Kucharski | | uucp: ...!{boulder,sun,uunet}!stan!kucharsk | Solbourne Computer, Inc. | = The opinions above are mine alone and NOT those of Solbourne Computer, Inc. =
gbrown@tybalt.caltech.edu (Glenn C. Brown) (03/17/90)
bill@bilver.UUCP (Bill Vermillion) writes: >In article <1990Mar15.090214.9871@spectre.ccsf.caltech.edu> gbrown@tybalt.caltech.edu (Glenn C. Brown) writes: > >> I'd be perfectly happy to settle for the telivision picture tubes of today! >>It's the signal that's so horrendous! >The output of a laser player isn't awesome. It's good... HA! I knew this discussion would get people going! My point was not that Laser Discs are awesome. My point IS that when *I* watch TV (not too often, mind you) It doesn't bother me that I can't get 6" from the sceen and still not see pixels: If I got a 6' screen, I would just sit 3 times further away from it than a 24" set. Sure, a few years after graduating from Caltech, I MIGHT be able to afford a HDTV set. But will it be WORTH it? Will I want to spend $5000 on a HDTV set, or would I rather spend $500 plop a box on the top of my TV that decodes digital broadcasts, and gives me a noise-free picture on my lower Res. monitor? I'd go for the $500 box! And what of the MILLIONS of people who won't be able to afford the HDTV sets? I say, there should be a middle-of-the-road solution in addition to HDTV. There should be HDTV, but I DON'T NEED HDTV! >I spent years in broadcast, and have seen many changes in the FCC, and I am >not too particularly impressed with their performance in the past few years. I can't argue w/ that! In fact, where does the FCC claim to get the legal authority to regulate speach over the airwaves? e.g. why will a HAM who says F**K on the airwaves almost surely lose his license? So much for freedom of speach! It's a form of government censorship. I have nothing against CENSURE, but censorship by the gov. is WRONG.
gbrown@tybalt.caltech.edu (Glenn C. Brown) (03/17/90)
bowers@elxsi.dfrf.nasa.gov (Al Bowers) writes: >>In article <1990Mar15.090214.9871@spectre.ccsf.caltech.edu> gbrown@tybalt.caltech.edu (Glenn C. Brown) writes: >Just as a point of reference laser disc has always been a superior to >tape. The fundamental reason for this is the digital nature of LDs. Uh, well, It's not exactly digital: It's PWM (pulse width modulation). This is the ANALOG recording of discrete samples. Therefore the recording is discrete, but I don't think it qualifies as digital. >I'd vote for a clean break from the current NTSC standard in the new >HDTV standard. Give me anywhere from 500 to 1000 lines horizontal >resolution and I could be happy with that until I'm old and gray... >Just another opinion... I hear you: GIVE ME. But are you willing to PAY for it? You sound like those comp.sys.mac people who say the low cost Mac should have at least a 68030 and 4 megs of RAM and 256 colors. :-P Glenn
fff@mplex.UUCP (Fred Fierling) (03/17/90)
In article <BOWERS.90Mar16105816@drynix.dfrf.nasa.gov>, bowers@elxsi.dfrf.nasa.gov (Al Bowers) writes: > > I'd vote for a clean break from the current NTSC standard in the new > HDTV standard. Give me anywhere from 500 to 1000 lines horizontal > resolution and I could be happy with that until I'm old and gray... This question comes up a lot doesn't it? Do you scrap the old technology and design a new, more refined and technically superior one, or do you compromise your design to maintain backwards compatibility? I hope the "clean break" approach wins out too. What a engineer's nightmare it would be to match up the side panels to the NTSC center. -- Fred Fierling uunet!van-bc!mplex!fff Tel: 604 875-1461 Fax: 604 875-9029 Microplex Systems Ltd 265 East 1st Avenue Vancouver, BC V5T 1A7, Canada
gbrown@tybalt.caltech.edu (Glenn C. Brown) (03/18/90)
fff@mplex.UUCP (Fred Fierling) writes: >I hope the "clean break" approach wins out too. What a engineer's nightmare >it would be to match up the side panels to the NTSC center. I heard that the FCC said they wouldn't approve the standard unless its backwards compatible...=-( Glenn
dave@imax.com (Dave Martindale) (03/18/90)
In article <sa0KhqO00Uh7M2R25C@andrew.cmu.edu> bas+@andrew.cmu.edu (Bruce Sherwood) writes: >At the risk of stating the obvious: > >Some of this discussion of HDTV implies that there will always be a need >for another generation of standards with even higher resolution. That >isn't necessarily the case. The human eye has limited resolution, and >higher resolution than that in the picture is literally useless, if you >are talking in terms of a "typical" screen size viewed from a "typical" >viewing distance. > >The analogy with audio is that a CD with frequency response out to 10 >MHz would not sound better than one with frequency response out to 20 >KHz, because the human ear can't hear the higher frequencies. If you move closer to a loudspeaker, you don't need better frequency response - your ear's limits are the same at any distance. You can make the same argument for colour and brightness resolution, but not spatial resolution, in an image. I.e. beyond a certain point, using extra bits for brightness or colour resolution just doesn't produce a noticeable improvement in the picture, no matter how close you get. But the analogy is all wrong for resolution. If you move closer to an image, so it fills more of your field of view, you need better spatial resolution. And somebody will always want to sit closer than the current standard is designed for, at least for the forseeable future. NTSC was designed for a viewing distance of 10 times the picture height, HDTV for 3-4 times the picture height. I want an image that looks sharp from 2/3 the picture height - that gives me a 90 degree field of view (with a 4:3 aspect ratio). What you say would be true only if there was a "typical" screen size and a "typical" viewing distance.
thant@horus.esd.sgi.com (Thant Tessman) (03/20/90)
In article <1990Mar17.022845.9450@spectre.ccsf.caltech.edu>, gbrown@tybalt.caltech.edu (Glenn C. Brown) writes: > > HA! I knew this discussion would get people going! > My point was not that Laser Discs are awesome. My point IS that > when *I* watch TV (not too often, mind you) It doesn't bother me that > I can't get 6" from the sceen and still not see pixels: [...] > > And what of the MILLIONS of people who won't be able to afford the > HDTV sets? The point of my original posting was that the FCC isn't even giving people the chance to choose. If new formats were allowed, the high fidelity nuts like me would be willing to support the HDTV industry until it became affordable for everyone, while leaving the old system in place until there was no longer enough of a market to support it. > > >I spent years in broadcast, and have seen many changes in the FCC, and I am > >not too particularly impressed with their performance in the past few years. > > I can't argue w/ that! In fact, where does the FCC claim to get the > legal authority to regulate speach over the airwaves? e.g. why will > a HAM who says F**K on the airwaves almost surely lose his license? So > much for freedom of speach! It's a form of government censorship. I > have nothing against CENSURE, but censorship by the gov. is WRONG. Wouldn't you consider regulating broadcast formats just as much censorship as regulating speech? thant
bowers@elxsi.dfrf.nasa.gov (Al Bowers) (03/20/90)
In article <1990Mar17.025325.9827@spectre.ccsf.caltech.edu> gbrown@tybalt.caltech.edu (Glenn C. Brown) writes: >bowers@elxsi.dfrf.nasa.gov (Al Bowers) writes: >>>In article <1990Mar15.090214.9871@spectre.ccsf.caltech.edu> gbrown@tybalt.caltech.edu (Glenn C. Brown) writes: ...story of my infamous analog digital screwup deleted... >>I'd vote for a clean break from the current NTSC standard in the new >>HDTV standard. Give me anywhere from 500 to 1000 lines horizontal >>resolution and I could be happy with that until I'm old and gray... >>Just another opinion... >I hear you: GIVE ME. But are you willing to PAY for it? You sound like >those comp.sys.mac people who say the low cost Mac should have at least >a 68030 and 4 megs of RAM and 256 colors. :-P Actually Glenn I have a real dislike for Macs and here is why, I refuse to use any machine whose presupposition is that I don't know anything about what it is I am trying to do. This prevents me from doing some of the wonderful things a Mac is capable of but it doesn't bother me too much. As to the implication that I would not be willing to spend the money to get the performance, I'd like to point out that if you'd read any of my previous postings you'd know that I firmly believe that you get what you pay for. My current camcorder is a S-VHS-C and next time around I will buy the same format with HiFi also. My VCR is a VHS and has MTS which came out in the mid '80s (MTS was only 12 months old when I bought it, only linear stereo but at the time that was all that was available). My next VCR will be the new JVC S-VHS(-C) machine that will accept S-VHS, VHS, S-VHS-C and VHS-C, it will cost me a pretty penny but it too will be worth it. GIVE ME the option and let me decide how much I want to spend for what features. :-P yourself... -- Albion H. Bowers bowers@elxsi.dfrf.nasa.gov ames!elxsi.dfrf.nasa.gov!bowers NASA Ames-Dryden Flight Research Facility, Edwards, CA Aerodynamics: The ONLY way to fly! Live to ski, ski to live...
sehat@iit (Sehat Sutardja) (03/20/90)
In article <1990Mar17.022845.9450@spectre.ccsf.caltech.edu>, gbrown@tybalt.caltech.edu (Glenn C. Brown) writes: > Sure, a few years after graduating from Caltech, I MIGHT be able to > afford a HDTV set. But will it be WORTH it? Will I want to spend > $5000 on a HDTV set, or would I rather spend $500 plop a box on the top > of my TV that decodes digital broadcasts, and gives me a noise-free > picture on my lower Res. monitor? I'd go for the $500 box! > Your wish might come true next year. For now, I can't say much about this. > And what of the MILLIONS of people who won't be able to afford the > HDTV sets? > This depends on how low the cost of a high resolution monitor would be. As far as the digital signal processing requirement goes, only a few milion transistors and some DRAM chips would be needed. You can pretty much guess what the cost of electronics would be when they are used in consumer products. For now, the major problem is to get a low cost monitor. -- Sehat Sutarja, {decwrl!sun}!imagen!iit!sehat | Integrated Information Tech. sutarja@janus.Berkeley.EDU | Santa Clara, CA. (408)-727-1885
minich@a.cs.okstate.edu (MINICH ROBERT JOHN) (03/21/90)
From article <sa0KhqO00Uh7M2R25C@andrew.cmu.edu>, by bas+@andrew.cmu.edu (Bruce Sherwood): > The analogy with audio is that a CD with frequency response out to 10 > MHz would not sound better than one with frequency response out to 20 > KHz, because the human ear can't hear the higher frequencies. Well, it probably would sound a bit better. Consider this: A 20KHz sample on CD looks something like this * * * * * * * * * * * * * * * * * * * * which is just a dumb square wave. Sure, it's at high enough of a pitch that most people wouldn't be able to discriminate between it and a pure sine of the same requency, but what happens if, say, you have a 20,001Hz waveform? Then, 20KHz just isn't enough to provide a nice, "symetric" waveform. Thus, you get a somewhat harsh sound. If I were really after a "human limits" sample, I'd bump the rate up to around 30KHz to minimize the distortion. (Assuming that a 40KHz sample is "wasteful".) Since there are indeed people sensitive enough to these at-the-limits conditions, we shouldn't write off any increase in the sapling freqeuncy as wasteful just because simple math says "you can't hear anything higher than..." The truth is, we CAN hear the effects BELOW the maximum frequency. Here's an analogy (and an excuse to post here): The human eye can only discern between a limited amount of colors, especially in small areas. The number is quite small (on the order of 100s). So, should we abandon 24bit color displays since we _shouldn't_ be able to tell the difference? Just because our hearing doesn't necessarily scream at us "yuck" like our eyes do doesn't mean we should ignore what does exist. I'll keep my 24bit color, thank you. And I'll also get the SUPER-CD player that sample at a higher rate because _I_ can tell the diff. Robert Minich Oklahoma State University minich@a.cs.okstate.edu DISCLAIMER: One who takes forgotten floppies.
keith@csli.Stanford.EDU (Keith Nishihara) (03/21/90)
minich@a.cs.okstate.edu (MINICH ROBERT JOHN) writes: >From article by bas+@andrew.cmu.edu (Bruce Sherwood): >> The analogy with audio is that a CD with frequency response out to 10 >> MHz would not sound better than one with frequency response out to 20 >> KHz, because the human ear can't hear the higher frequencies. > Well, it probably would sound a bit better. Consider this: >A 20KHz sample on CD looks something like this > * * * * * * * * * * >* * * * * * * * * * >which is just a dumb square wave. Sure, it's at high enough of a pitch >that most people wouldn't be able to discriminate between it and a pure sine I can't take this any more! You don't just feed the samples through an audio amlifier and see the square wave! You put them through a `reconstruction filter' which reconstructs the waveform. An ideal reconstruction filter, with a step function low pass frequency response at 20kHz will reconstruct the 20kHz waveform as a *perfect* sine wave. It will also reconstruct a 19.99 kHz waveform *perfectly*, notwithstanding the fact that there is a beat between the sample frequency and the frequency represented (the sample points `walk' slowly along the wave shape). So if the basilar membrane in your ear responds up to 20kHz you _will not hear_ the difference between a properly reconstructed signal from 44.1kHz samples and a signal reconstructed from 20MHz samples! Most adults' hearing is far below this limit, in any case (15kHz is considered good -- if you have often: operated heavy machinery, fired a gun, driven a car with the window open, or listened to loud music with headphones on, 8 to 12 kHz may be more like it!) Before someone asks what if the original were not a sine wave: recall that complex waveforms may be considered as a summation of sine waveforms of different amplitudes and frequencies, so in a linear system it is valid to think only in terms of the behaviour of the individual sine wave components. Of course, perfect reconstruction filters are hard build, so a 44.1 kHz sample rate permits reconstruction filters to have a finite roll off starting at 20kHz and being essentially fully cut at 22.05kHz (the limit for a 44.1 kHz filter), and *still* repro- duce all frequencies up to 20kHz *perfectly*. Now if the filter did not cut off frequencies above 22.05 kHz, some of that 20 kHz signal would appear as a 24.1 kHz signal (reflected in frequency about the Nyquist frequency). This would be undesirable. Oversampling (no one sells CD players that don't `oversample' any longer, do they?) permits some of the reconstruction filtering to be done using a digital filter. Consider 4x resampling: each sample is replicated four times in a row at 176.4 kHz. A digital filter with a cut off frequency of 20 kHz can be applied. Now when reconstructing, the analog filter still has to be flat to 20kHz, but need not be fully cut until 88.2 kHz, the Nyquist rate for the 4x oversampled signal. Since the digital filter has en- sured that there will be no frequency components in the digital signal between 20kHz and 88.2kHz, a much lower Q filter may be used, which is much easier and cheaper to design. Now what about those 18 bit players? CDs only have 16 bit sam- ples dont they? but if you use oversampling and digital filter- ing, you can `interpolate' between the original samples and sam- ple quantisation. But what does it buy you? The reconstructed signal is only as good as the orignal digital material. A good advertising gimmick, in my opinion. (What about the precision and linearity of those 18 bit A-D converters?) Neil/. Neil%teleos.com@ai.sri.com Note that our mail feed via SRI is currently dead, so that flames, questions and assertions that `my hearing is good to 37.496e29 MhZ -- medically verified' (you must be an alien) will be thrown into the bit bucket.
turk@media-lab.media.mit.edu (Matthew Turk) (03/22/90)
In article <5478@okstate.UUCP> minich@a.cs.okstate.edu (MINICH ROBERT JOHN) writes: > > Well, it probably would sound a bit better. Consider this: > > A 20KHz sample on CD looks something like this > > * * * * * * * * * * > * * * * * * * * * * > > which is just a dumb square wave. Sure, it's at high enough of a pitch > that most people wouldn't be able to discriminate between it and a pure sine > of the same requency, but what happens if, say, you have a 20,001Hz waveform? > Then, 20KHz just isn't enough to provide a nice, "symetric" waveform. Thus, > you get a somewhat harsh sound. If I were really after a "human limits" sample, > I'd bump the rate up to around 30KHz to minimize the distortion. (Assuming that > a 40KHz sample is "wasteful".) ... > The truth is, we CAN hear the effects BELOW the maximum frequency. The problem is quite a bit better understood than you are assuming. The aliasing you describe is eliminated by prefiltering the signal with a lowpass filter. Also, digital signals are not reproduced as square waves. If the human ear was indeed insensitive to signals above 20kHz, then an ideal system would prefilter the signal (lowpass at 20kHz), sample at 40kHz, then reconstruct the (filtered) analog signal exactly to be amplified and sent to your speakers. The real issues here are: (1) a perfect low-pass filter is not realizable, so you either have to accept some aliasing or filter at a higher rate; (2) the human frequency response isn't an ideal low-pass system, so there's no clear and clean cutoff point. ~20kHz is, I believe, the -3dB point. Since the CD sampling rate is 44.1kHz, there's a little room for variation -- perfect filtering would fully represent signals < 22.05kHz. In real systems, we can definitely avoid any noticable aliasing, but this reduces the frequence response. Anyone know how tight the filters used in digital recording are? Matthew
bowers@elxsi.dfrf.nasa.gov (Al Bowers) (03/22/90)
In article <5478@okstate.UUCP> minich@a.cs.okstate.edu (MINICH ROBERT JOHN) writes: >From article <sa0KhqO00Uh7M2R25C@andrew.cmu.edu>, by bas+@andrew.cmu.edu (Bruce Sherwood): >> The analogy with audio is that a CD with frequency response out to 10 >> MHz would not sound better than one with frequency response out to 20 >> KHz, because the human ear can't hear the higher frequencies. > Well, it probably would sound a bit better. Consider this: >A 20KHz sample on CD looks something like this > * * * * * * * * * * >* * * * * * * * * * Exactly! Or maybe I would have described it as: _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ >which is just a dumb square wave. Sure, it's at high enough of a pitch >that most people wouldn't be able to discriminate between it and a pure sine >of the same requency, but what happens if, say, you have a 20,001Hz waveform? >Then, 20KHz just isn't enough to provide a nice, "symetric" waveform. Thus, >you get a somewhat harsh sound. If I were really after "human limits" sample, >I'd bump the rate up to around 30KHz to minimize the distortion.(Assuming that >a 40KHz sample is "wasteful".) I realize that this is a little out of place but as an example we in the aircraft industry prefer to sample at a minimum of 5 times the maximum frequency of interest and we usually prefer 10 times the max frequency. Now I realize that this is far and away more than required for decent sound (or maybe even exceptional sound) but having the cut off just above the limit of average human hearing just seems to be a little rash. -- Albion H. Bowers bowers@elxsi.dfrf.nasa.gov ames!elxsi.dfrf.nasa.gov!bowers NASA Ames-Dryden Flight Research Facility, Edwards, CA Aerodynamics: The ONLY way to fly! Live to ski, ski to live...
kassover@jupiter.crd.ge.com (David Kassover) (03/22/90)
In article <BOWERS.90Mar21114921@drynix.dfrf.nasa.gov> bowers@elxsi.dfrf.nasa.gov (Al Bowers) writes: ... >I realize that this is a little out of place but as an example we in >the aircraft industry prefer to sample at a minimum of 5 times the >maximum frequency of interest and we usually prefer 10 times the max >frequency. ... I also work in conjunction with the aircraft industry, and have a devil of a time convincing the airframe and powerplant types that sampling at more than their beloved 10x maximum frequency just introduces sampling noise into the analysis (and therefore anything downstream of the sampler) (As well as making it hard for me to build hardware and software that can actually sample *and* do the required calculations that fast 8-) ) Remembering way back into High School Health, the generally accepted *nominal* range of human hearing is 20-20kHz. I don't think there's a problem with building audio components that are "flat" (well, flat enough. +- .5 dB?) out to say 30kHz. And economic for most of us to buy. If someone wants to spend more than that and get "better" frequency response, they may. Assuming, of course, that the input signal contains meaningful information at those high frequencies, anyway. (NOT whether it should, just whether it does). The case of an audio frequency square wave is somewhat misleading, since the mechanical components of the system possess enough inertia to low-pass filter the signal. (like speaker cones, eardrums, and ossicles) Back to the point. The standard that is adopted should be capable of providing a *reasonable* benefit for most people. Even unreasonably, say 3.5 sigma out from "normal" or "average", and furthermore should allow for those 4.0+ sigma people to add on, at additional cost to themselves. I don't see why there is so much agony about making the people who want video to do traditional graphics and the people who want video to look at more true-to-life images have to put up with the same standards. I submit that *most* people who do both kinds of activities will not need to convert from one to the other often, and those who do will be able to get hardware, software, or whatever to do it. I submit that it is not necessary to displease everyone, nor to provide the new technology immediately to the man on the street for only $49.95 in 1990 dollars. Before you all flame me for wasting bandwidth: I am reading this through comp.graphics. The article I am following-up was posted to no less than 3 other newsgroups, none of which I read. IMHO, the proposed standard for HDTV is a reasonable piece of information for posting here. The argument, polemics, and other such stuff is maybe better restricted to someplace else.
johna@gold.GVG.TEK.COM (John Abt) (03/22/90)
In article <5478@okstate.UUCP> minich@a.cs.okstate.edu (MINICH ROBERT JOHN) writes: >From article <sa0KhqO00Uh7M2R25C@andrew.cmu.edu>, by bas+@andrew.cmu.edu (Bruce Sherwood): >> The analogy with audio is that a CD with frequency response out to 10 >> MHz would not sound better than one with frequency response out to 20 >> KHz, because the human ear can't hear the higher frequencies. > > Well, it probably would sound a bit better. Consider this: >A 20KHz sample on CD looks something like this > * * * * * * * * * * >* * * * * * * * * * >which is just a dumb square wave. Sure, it's at high enough of a pitch >that most people wouldn't be able to discriminate between it and a pure sine >of the same requency, but what happens if, say, you have a 20,001Hz waveform? >Then, 20KHz just isn't enough to provide a nice, "symetric" waveform. Thus, >you get a somewhat harsh sound. But the 20 KHz square wave is just a "dumb" sine wave after it goes through the re-construction filter. And, as any student of Fourier will tell you, the only thing that can change the periodic shape of a sine wave are harmonics - the first of which for a 20 KHz waveform occurs at 40 KHz. Nobody can hear 40 KHz. >...... Here's >an analogy (and an excuse to post here): The human eye can only discern between >a limited amount of colors, especially in small areas. The number is quite >small (on the order of 100s). So, should we abandon 24bit color displays since >we _shouldn't_ be able to tell the difference? Bad analogy because it's not always applicable, e.g., the eye is extremely sensitive to correlated discontinuities in an image. The number of different colors that are discernable when seperated by distinct line is far greater. Matter of fact, 10 bit RGB makes better pictures. - John Abt, Grass Valley Group